The Ledger
Every statement that passed quotation and attribution checks. Mix any filter with any other: certainty 1/5, debate potential 5/5, or both at once.
why aren't all 23 resolved? a statement only gets an assessment when the public record can support or contradict it. opinions and what-ifs never can, and 0 checkable ones are still open, waiting for their date. predictions held up or didn't; assertions are supported or contradicted. on every card: ▮▮▮▮▮ certainty · ▮▮▮▮▮ debate potential. speakers are clickable
Delangue: No Single AI Model Will Dominate Because AI Resembles Codebases
“In the simplistic terms, the model or like AI is like a code base, right? It's a bit different, but at the end of the day, it's a code base, right? And so it's kind of silly to say, oh, this code base is better than this code base, or there's going to be one c…”
Delangue: Using AI APIs Is Like Building Websites on Squarespace
“The analogy that I sometimes like is that on the early days of the web, you could use kind of like something that would Create a website for you, right? You could use the equivalent of a Squarespace or for weeks, and that would make you feel good, right? Becau…”
Delangue: AI Is Far From Becoming Autonomous or Humanity-Destroying
“I believe at the same time, we are very far from a world where AI is autonomous. And has conscience and is taking over the world and destroying humanity. I think this is more fear that is very sci-fi driven that we're very far from.”
Delangue: 95% of a VC's Value Is Financial Capitalization and Support
“Investors are first and foremost investors, right? Meaning that their main value add is to deal around To help you on financial matters. So for example, when SVB go down, right? And then to help you to always capitalize the company the right way. So they're, i…”
Delangue: VCs Should Not Act Like Operators or CEOs
“They sometimes act almost as CEO or, like, operators for companies. Which in, in my opinion is, is not really their job. And worse than that, I feel like sometimes entrepreneurs are building companies for investors and investors are behaving like entrepreneurs…”
Clément Delangue: Deep AI Startups Can Execute 100x Better Than Incumbents
“I think if you were thinking about AI as AI APIs, I agree. If you're thinking of AI as a more radical paradigm switch to build technology, right? And if you think about an AI startup as a company that is actually training models, creating new architectures, op…”
Delangue: Startups training their own models build better products than wrappers
“If you look at models closer to Runway ML or Stability AI or like PhotoRoom in Paris, you see these AI native startups that are actually building, training their own models. And how they can, in my opinion, build much better things than the ones that just use …”
Delangue: Open Source Accelerated AI Progress by 30 to 50 Years
“Without open science, without open source, without Google sharing their attention is all you need paper, sharing their birth paper the latent diffusion paper. Maybe we would be 3040, 50 years away from where we are today.”
Delangue: Incumbents Struggle With AI Due to Long-Term Research Requirements
“Because it's a completely different way to build technology, right? It's a way where you have to have scientists, for example, working for six months on a new architecture, on a new model before releasing it. So it's a different paradigm of how you build softw…”
Delangue: The ROI on Training Larger AI Models Is Declining
“We starting to realize that more compute for models is not necessarily The right thing, or at least that it's not enough and the return on investment on training larger and larger models is starting to go down.”
Delangue: Hugging Face Refuses to Speak With External VCs Between Rounds
“One of these rules is that I don't talk to any investors external investors in between rounds, right? I'm making an exception for you today because it's a podcast, but Otherwise, I don't talk to anybody in between rounds because I feel like a lot of time it's …”
Delangue: Every Company Will Eventually Build and Run Proprietary AI Models
“I think the biggest thing is, in my opinion, all companies will have their own AI models. Like all companies will have their chat GPT or their GPT-IV.”
Delangue: Only 50 to 100 People Globally Can Train State-of-the-Art AI
“I would say, you know, machine learning engineer, and by machine learning engineer, I mean someone who's really building a new architecture for AI models and able to train state of the art models. They are just in my opinion, a few people in the world who has …”
Delangue: VC Interest in AI Lags Behind Years of Massive Usage
“The VC and the mainstream interest is kind of like a catch up on the reality. Because if you look at usage of AI, it's been massive and it's been growing massively for the past, you know, three years, right? Even before, before ChatGPT, even before like, you k…”
Delangue: 10 of 13 Meta LLaMA Authors Are Based in Paris
“I don't have the exact number, but I think 10 out of the 13 authors of Lama are actually based in, in Paris, right? In, in the Meta AI, lab that, that they have there that is huge.”
Delangue: AI Industry Will Gain Fair Use Regulatory Clarity in 2023
“And it's gonna be good this year, I think, because we're going to start to have more legal clarity about, you know, how do we consider fair use what are the regulators expecting from AI companies to respect in terms of rules.”
Delangue: Hugging Face Was First to Train AI on Opt-Out Datasets
“I think we were the first organization to train on a fully opted out data sets where developers could just remove themselves from the training.”
Hugging Face Has 15,000 Corporate Users and 3,000 Paying Customers
“We have 15,000 companies using us now. And then a smaller subset of companies are actually paying us, right? And for us, it's 3000 companies.”
Meta, Bloomberg, and Grammarly Are Paying Enterprise Customers of Hugging Face
“3000 companies are paying us for that, including Meta, including Bloomberg, including Grammarly and companies like that.”
Delangue: Hugging Face's Monetization Model Will Shift Within Five Years
“Probably the way we make money today is not going to be the way we make money in three years or in five years.”
Delangue: AI Startups Cost More to Build Than Traditional Software Startups
“It does, a little bit. A bit similarly, in my opinion, to how, you know, like you would build like an internet company or like a software company, 20 years ago. Because compute is kind of like more expensive for AI than it is for traditional software. Because …”
Delangue: Fundraising Difficulty Depends on Momentum, Not Company Stage
“Something I didn't expect is that the way you raise Isn't so much dictated by your stage and your rounds. That's more dictated by the situation that you're in, in terms of traction, in terms of momentum, in terms of like achievements, right?”